Until recently, the majority of published or publicly available solid reconstruction models lacked an integral three-dimensional topographical context for the modelled structures. Buildings stand in isolation on a flat featureless plain, rather than in a representation of their surrounding three-dimensional space. Where topographic representations are used, these tend to be presented as theatrical backdrops to the real object of study the modelled structures themselves. An early exception is the reconstruction model of a motte and bailey castle and later medieval enclosure at Mathrafal, Wales ( ), which incorporated a topographic survey with interpretative models, but the reconstruction of the actual structures was very primitive.

With increased emphasis on the generation of 'reality', the context of the models assumes more importance. More recent models tend apply a range of different techniques. For example, reconstructions based around San Vincenzo al Volturno <http://www.digis.it/sanvin/> use a photographic backdrop to contextualise the architectural model of the abbey and there is also a limited local generalised topography provided in some models.

Increasingly, there is an interface between interpretative structural modelling and Geographical Information Systems, with topographic data collected with GIS applications in mind being incorporated within the 3-D reconstruction model. Classic examples of this are the models of Peel Gap and Hepburn bastle house by Gillings and Goodrick ( ). A micro-topographic survey of the area surrounding the Peel Gap tower on Hadrian's Wall is incorporated within a VRML model in order to assess visibility. The model of Hepburn bastle uses a combination of Ordnance Survey map data and EDM-derived data to provide a limited model of the landscape context of the structure. More elaborate examples include the new 'virtual' Stonehenge which employs data derived from a 6km2 survey of the surrounding landscape ( ) and the Giza Plateau model ( ) which incorporates a detailed terrain model of a 3.54km area of the Giza plateau with a range of modelled pyramids, mortuary and valley temples and around 150 mastaba tombs.

However, in most of these cases the topography is still often little more than a theatrical backdrop for the structures, except where it has been used in visibility studies, for example. It provides a stage for viewing the model, but is hardly integral to the model itself. There are no instances in which the topography is the model where topographical details accompanied by structural elements are integral to the model and what constitute the model. The most obvious situation in which this would arise is with reconstruction models of earthworks with structures erected upon them, but, Mathrafal aside, such models are largely unknown. Surface models of earthworks much like those in Figures are used, but not incorporated with interpretative structural reconstructions. A recent model of the Danebury hillfort entrance comes close this consists of a surface model constructed to assess visibility ( ), but, curiously, in the light of this omits any attempt to represent palisades or other structural elements. Why these models are so rare is again unclear, but it may be significant that they would more often than not involve the reconstruction of timber structures.

URL: http://intarch.ac.uk/journal/issue8/huggett/jhissues3dc.html